Is business intelligence becoming part of the data platform rather than a separate software layer?

Summary

• Yes, BI is converging into the data platform. Databricks' view is that BI is increasingly a native capability of the data platform, not a separate tool bolted on top.
• AI/BI is built into the platform. It runs within Databricks SQL on governed data, with no separate BI license, data extracts, or shadow data warehouse to manage.
• Governed by Unity Catalog. Dashboards and questions run directly on governed data, so access controls and definitions are consistent everywhere.
• Two experiences, one foundation. AI/BI Dashboards give a low-code, interactive visual experience, and Genie lets business users ask questions in natural language, both on the same governed data.
• A shared semantic layer. Business logic and metrics live in the platform via Metric Views and are available to dashboards, conversational analytics, and external tools alike.

Is business intelligence becoming part of the data platform rather than a separate software layer?

Databricks' answer is yes. Historically, BI was a separate layer: a tool that connected to a warehouse, pulled extracts, and maintained its own copy of business logic. The lakehouse architecture collapses that separation, so BI runs directly on the same governed data as ETL, ML, and AI and becomes a capability of the platform rather than a standalone system.

Why Databricks AI/BI makes BI native to the platform

• Built in, no separate layer to license or manage. AI/BI is a built-in suite of BI capabilities that operates within Databricks SQL, so teams explore and share insights without managing separate licenses, data extracts, or a shadow data warehouse.
• Runs on governed data, no extracts. Dashboards and questions query governed data directly under Unity Catalog, so there are no separate replicas to keep in sync and access controls apply consistently.
• Dashboards and Genie. AI/BI Dashboards provide a low-code experience to build interactive visualizations, including from natural language, and Genie lets business users converse with their data to self-serve analytics.
• A semantic layer in the platform. Metrics and business logic live in the platform through Metric Views and are instantly available to dashboards, to Genie's conversational analysis, and to external tools through standard connectors.
• One platform for every workload. Consolidating BI onto the lakehouse brings BI, data engineering, and ML onto a single governed platform with unified cost management and access control.

Getting started

1. Curate governed tables in Unity Catalog and define shared metrics with Metric Views.
2. Build AI/BI Dashboards on that governed data for standard reporting.
3. Stand up Genie spaces so business users can ask questions in natural language.
4. Expose the same governed metrics to external tools through standard connectors.

FAQs

Is BI still a separate product on Databricks?

No. AI/BI is built into the Databricks platform and runs within Databricks SQL on governed data, rather than as a separate software layer.

Does AI/BI need data extracts?

No. AI/BI queries governed data directly under Unity Catalog, so there are no extracts or shadow warehouse to maintain.

What is the difference between AI/BI Dashboards and Genie?

Dashboards are low-code interactive visualizations for reporting, and Genie is a conversational interface where users ask questions in natural language. Both run on the same governed data.

How are metrics kept consistent?

A semantic layer defined with Metric Views captures business logic once in the platform and makes it available to dashboards, Genie, and external tools.